Brain injury is the main cause auf long-term disability in children and carries the greatest disease burden out of any pediatric condition. Children with brain injury often compensate remarkably well and can develop abilities typically supported by brain structures now damaged. It is thought that much of this recovery results from the use-driven reorganization of the brain's complex functional network architecture. A better understanding of the links between activity, network organization and functional recovery is paramount to developing much needed novel neurorehabilitative treatments and improving existing ones. Therefore, it is our long-term objective to uncover the systems-level fundamentals of use-driven functional network plasticity in children with brain injury. Constraint-induced movement therapy (CIMT), a treatment for hemiparesis, provides an excellent research model for studying use-driven brain reorganization. Patients treated with CIMT have their stronger upper extremity restrained in a cast for three weeks while practicing difficult tasks with the affected side. CIMT radically alters upper extremity use, thus lastingly improving function on the treated side. The objective of the current proposal is to learn how constraint-induced movement therapy (CIMT) improves motor function in children with chronic brain injury. The specific aims are to 1) identify those functional network changes most important for improved upper extremity use, 2) establish neuropsychometric markers that predict treatment response, and 3) investigate the effect of therapy dose on outcomes. This study will test our hypothesis that much of CIMTs benefit derives from a reweighting of functional connections between the brain's attention and motor networks towards the treated side and that these changes are modulated by attentional dysfunction and therapy dose. To achieve these aims we will repeatedly assess functional networks and motor behavior with multi-modal MRI (functional, functional connectivity) and wearable accelerometer biosensors, in children undergoing CIMT. A within-subject design and novel single-subject network analytics will for the first time allow us to factor out lesion inhomogeneity and extract commonalities of therapy-driven changes across patients. This study will strongly impact health-related research by identifying beneficial connections that should be targeted with brain stimulation and neurofeedback and potentially revealing modifiable patient (attentional dysfunction) and therapy factors (dose) critical to therapy success. Candidate: Dr. Nico Dosenbach, the candidate, is a systems neuroscientist and pediatric neurologist whose prior research using functional MRI (fMRI) and functional connectivity MRI (fcMRI) has contributed significantly to the understanding of attentional control networks and functional network development. Driven by his clinical interest in advancing rehabilitative treatments for pediatric brain injury, Dr. Dosenbach is seeking additional training in movement science, developmental neuropsychology, biostatistics and clinical pediatric neurorehabilitation. Dr. Dosenbach's career goal is to improve the scientific understanding of functional network reorganization in pediatric brain injury in order to develop novel neurorehabilitative treatments, as well as to optimize current treatments. His career development plan includes training in movement science with Dr. Lang, developmental neuropsychology with Dr. Barch, biostatistics with Dr. Shannon and in clinical pediatric neurorehabilitation with Dr. Noetzel. Dr. Dosenbach's principal mentor, Dr. Bradley Schlaggar, a pediatric neurologist and developmental cognitive neuroscientist, will also provide training in movement science and developmental neuropsychology and convey the qualities and strategies of a successful, independent physician-scientist. Dr. Dosenbach will complete classes in movement science, neuropsychology, biostatistics and Research Ethics. Environment: Washington University's (WU) neuroimaging research community is one of the largest and most highly regarded and has a reputation for openness and strongly supporting young investigators. Many of the leading neuroscientists studying human functional networks, such as Drs. Marcus Raichle, Steven Petersen, Maurizio Corbetta and Bradley Schlaggar, work at WU. The Human Connectome Project (HCP) led by Dr. David Van Essen and Dr. Deanna Barch has brought even greater neuroimaging resources and expertise to WU. In addition, WU is nationally highly ranked (US News and World Report) in pediatric neurology (# 5), occupational therapy (#2) and physical therapy (#3). Overall, WU is the single best institution imaginable for Dr. Dosenbach's research career development and the success of the proposed project.